A hybrid approach to NER by MEMM and manual rules

Moshe Fresko*, Binyamin Rosenfeld, Ronen Feldman

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

14 Scopus citations

Abstract

This paper describes a framework for defining domain specific Feature Functions in a user friendly form to be used in a Maximum Entropy Markov Model (MEMM) for the Named Entity Recognition (NER) task. Our system called MERGE allows defining general Feature Function Templates, as well as Linguistic Rules incorporated into the classifier. The simple way of translating these rules into specific feature functions are shown. We show that MERGE can perform better from both purely machine learning based systems and purely-knowledge based approaches by some small expert interaction of rule-tuning.

Original languageAmerican English
Title of host publicationCIKM'05 - Proceedings of the 14th ACM International Conference on Information and Knowledge Management
Pages361-362
Number of pages2
DOIs
StatePublished - 2005
Externally publishedYes
EventCIKM'05 - Proceedings of the 14th ACM International Conference on Information and Knowledge Management - Bremen, Germany
Duration: 31 Oct 20055 Nov 2005

Publication series

NameInternational Conference on Information and Knowledge Management, Proceedings

Conference

ConferenceCIKM'05 - Proceedings of the 14th ACM International Conference on Information and Knowledge Management
Country/TerritoryGermany
CityBremen
Period31/10/055/11/05

Keywords

  • Information Extraction
  • Machine Learning
  • Maximum Entropy Markov Model
  • Named Entity Recognition
  • Text Mining

Fingerprint

Dive into the research topics of 'A hybrid approach to NER by MEMM and manual rules'. Together they form a unique fingerprint.

Cite this